DE. J. P. JOULE ON THE SUEFACE-CONDENSATION OF STEAM. 
137 
On the other hand, the small quantity of water drawn off at W at the termination of an 
experiment is always more or less heated. Corrections on both these accounts were 
easily applied. 
I had at first some doubts whether the vacuum would not become gradually impaired 
by ah’ coming over from the boiler ; for it has been frequently asserted that water be- 
comes perfectly free from air only after long-continued boiling. I found, however, that 
after boiling had taken place for only two or three minutes, the air was entirely expelled, 
and that even if condensation were afterwards carried on until the receiver was entirely 
filled with water, no change took place in the height of the gauge. Hence, by blowing 
off steam for ten minutes at the commencement of a day’s experimenting, I effectually 
secured myself against any risk of the interference of air*. 
The Table of experiments requires little explanation. It will be seen that column 5 
contains some numbers with the negative sign. This might be expected where a small 
quantity of water was used, on account of its being raised in temperature during its 
ascent. When the water was intended to go in the same direction as the steam, it was 
pom’ed in at the upper end of the outer tube, and flowed away at the lower end, the 
pipe E being removed. Each number in the 14th column is the average of all the 
obseiTations of the pressure in the condenser after it became constant; and column 
17 contains the averages of all the observations of the temperature of the refrigerating 
water at its overflow made at the moments of gauge-observation. Hence this column 
contains numbers generally a little different from those of column 7, which, being taken 
for the purpose of deducing the total heat of steam, are the averages of all the tem- 
perature observations of the overflow water in the several experiments. 
In order to explain the principle on which the 18th column is based, I cannot do 
better than give textually the extract of a letter I received from Professor Thomsox, to 
whom at the outset I communicated my design, and who, with his usual zeal and kind- 
ness, immediately offered me very valuable suggestions. 
“ Steamer Venus, August lOtli, 1859. 
“ If the resistance to equalization of temperature between the steam and water de- 
pended on conduction through the separating metal alone, the heating effect would take 
place according to the law you name. The formula would be thus found, 
, , Xdx 
ivdv= —k V. 
a ’ 
where w is the mass of water passing per unit of time, dv the augmentation of the 
* I could not discover any alteration in tlie composition of the air after it had remained in the boiler some 
days. There appears to be no truth in the hypothesis which ascribes boiler explosions to the formation of 
hydrogen. The obvious cause is over-pressure ; and it is not wonderful that, when multitudes of boilers are 
worked at a very considerable proportion of the pressure calculated to burst them when new, accidents 
occasionally occur. I have repeatedly insisted upon the absolute necessity of periodical testing, and have 
proposed a method requiring no extra apparatus or expense, which consists simply in lighting a fire under 
the boiler when completely filled, and so producing the proof pressure by the expansion of w'ater by heat. 
I try my boiler every six weeks by this process, which appears to answer the end in view in every respect. 
